from sklearn.model_selection import KFold, cross_val_score, train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor
from sklearn.metrics import r2_score
import pandas as pd
import joblib
# Veri setini yükle
data = pd.read_csv("musicdata_standardized.csv")
# Bağımsız değişkenler ve hedef değişkenleri ayır
X = data.drop(columns=["Artist Popularity", "Track Popularity"])
y_artist = data["Artist Popularity"]
y_track = data["Track Popularity"]
# Veri setini %80 eğitim, %20 test olarak ayırma
X_train_artist, X_test_artist, y_train_artist, y_test_artist = train_test_split(X, y_artist, test_size=0.2, random_state=47)
X_train_track, X_test_track, y_train_track, y_test_track = train_test_split(X, y_track, test_size=0.2, random_state=47)
# Model tanımları
models = {
"Linear Regression": LinearRegression(),
"Decision Tree Regression": DecisionTreeRegressor(),
"Random Forest Regression": RandomForestRegressor(),
"Gradient Boosting Regression": GradientBoostingRegressor()
}
best_model_name_artist = None
best_score_artist = -1
# Artist Popularity için model seçimi ve performans ölçümü
print("Artist Popularity için Modellerin Değerlendirilmesi:")
print("--------------------------------------")
for name, model in models.items():
# Çapraz doğrulama ile modeli değerlendir
cv_scores_artist = cross_val_score(model, X_train_artist, y_train_artist, cv=10, scoring='r2')
mean_score_artist = cv_scores_artist.mean()
if mean_score_artist > best_score_artist:
best_score_artist = mean_score_artist
best_model_name_artist = name
# Modeli eğitim seti üzerinde eğit
model.fit(X_train_artist, y_train_artist)
# Test R^2 hesapla
test_predictions_artist = model.predict(X_test_artist)
test_r2_artist = r2_score(y_test_artist, test_predictions_artist)
# Eğitim R^2 hesapla
train_predictions_artist = model.predict(X_train_artist)
train_r2_artist = r2_score(y_train_artist, train_predictions_artist)
# Sonuçları yazdır
print(f"{name}: Cross-validated R^2 scores: {cv_scores_artist}")
print(f"{name}: Average Cross-validated R^2 score: {mean_score_artist}")
print(f"{name}: Test R^2 score = {test_r2_artist}")
print(f"{name}: Train R^2 score = {train_r2_artist}")
print("--------------------------------------")
# En iyi modeli seç (Artist Popularity için)
best_model_artist = models[best_model_name_artist]
# Track Popularity için model seçimi ve performans ölçümü
best_model_name_track = None
best_score_track = -1
print("\nTrack Popularity için Modellerin Değerlendirilmesi:")
print("--------------------------------------")
for name, model in models.items():
# Çapraz doğrulama ile modeli değerlendir
cv_scores_track = cross_val_score(model, X_train_track, y_train_track, cv=10, scoring='r2')
mean_score_track = cv_scores_track.mean()
if mean_score_track > best_score_track:
best_score_track = mean_score_track
best_model_name_track = name
# Modeli eğitim seti üzerinde eğit
model.fit(X_train_track, y_train_track)
# Test R^2 hesapla
test_predictions_track = model.predict(X_test_track)
test_r2_track = r2_score(y_test_track, test_predictions_track)
# Eğitim R^2 hesapla
train_predictions_track = model.predict(X_train_track)
train_r2_track = r2_score(y_train_track, train_predictions_track)
# Sonuçları yazdır
print(f"{name}: Cross-validated R^2 scores: {cv_scores_track}")
print(f"{name}: Average Cross-validated R^2 score: {mean_score_track}")
print(f"{name}: Test R^2 score = {test_r2_track}")
print(f"{name}: Train R^2 score = {train_r2_track}")
print("--------------------------------------")
# En iyi modeli seç (Track Popularity için)
best_model_track = models[best_model_name_track]
# Artist Popularity için en iyi modelin performansını hesapla
best_model_artist.fit(X_train_artist, y_train_artist)
train_predictions_artist = best_model_artist.predict(X_train_artist)
train_r2_best_artist = r2_score(y_train_artist, train_predictions_artist)
test_predictions_artist = best_model_artist.predict(X_test_artist)
test_r2_best_artist = r2_score(y_test_artist, test_predictions_artist)
# Track Popularity için en iyi modelin performansını hesapla
best_model_track.fit(X_train_track, y_train_track)
train_predictions_track = best_model_track.predict(X_train_track)
train_r2_best_track = r2_score(y_train_track, train_predictions_track)
test_predictions_track = best_model_track.predict(X_test_track)
test_r2_best_track = r2_score(y_test_track, test_predictions_track)
# En iyi modelin adını ve performansını yazdır (Artist Popularity)
print(f"\nEn iyi model (Artist Popularity): {best_model_name_artist}, Ortalama çapraz doğrulama R^2: {best_score_artist}")
print(f"En iyi model ({best_model_name_artist}): Eğitim R^2 skoru = {train_r2_best_artist}, Test R^2 skoru = {test_r2_best_artist}")
# En iyi modelin adını ve performansını yazdır (Track Popularity)
print(f"\nEn iyi model (Track Popularity): {best_model_name_track}, Ortalama çapraz doğrulama R^2: {best_score_track}")
print(f"En iyi model ({best_model_name_track}): Eğitim R^2 skoru = {train_r2_best_track}, Test R^2 skoru = {test_r2_best_track}")
# En iyi modeli diske kaydet
joblib.dump(best_model_artist, 'Best_Model_Artist_Popularity.pkl')
joblib.dump(best_model_track, 'Best_Model_Track_Popularity.pkl')
